The Unassuming Economist

From VOX post by Peter Hooper, Frederic S. Mishkin, Amir Sufi:

“The apparent flattening of the Phillips curve has led some to claim that it is dead. The column uses data from US states and metropolitan areas to suggest a steeper slope, with non-linearities in tight labour markets. We have been here before – in the 1960s, similar low and stable inflation expectations led to the great inflation of the 1970s. 

At a ‘Fed Listens’ event on 26 September 2019, Richard Clarida, vice chair of the Federal Reserve Board, observed that the flattening of the Phillips curve in recent decades is central to the Fed’s review of policy strategy (Clarida 2019). Price inflation has become much less responsive to resource slack, permitting the Fed to support employment during economic downturns more aggressively than it has in the past.

A flat Phillips curve reduces the chances of a breakout of inflation. This is especially important because the Fed considers the benefits of running a high-pressure economy, and of adopting a policy strategy that makes up for inflation misses to the downside by aiming for subsequent overshoots. Many participants in financial markets go even further than the Fed, believing that the Phillips curve is dead – in other words, excessive inflation is no longer a risk.

Recent experience in the US, Europe, and Japan appears to support this view. Major central banks struggle to get inflation to return to (or even move towards their objectives), even after labour markets have tightened. The US labour market has been running at or beyond estimates of full employment for the past two years, and inflation is still significantly below the Fed’s 2% target.

Indeed, measures of inflation expectations have been drifting lower, not higher as the Phillips curve model would predict. So is this model really dead, or just dormant? If not dead, how can we explain the flattening of the Phillips curve? What might reverse this trend, leading to a resurgence of inflation?

A lot of empirical research has been devoted to these questions over the past decade, for example Yellen (2015), Kiley (2015) Blanchard (2016), Nalewaik (2016), Powell (2018), and Hooper et al. (2019). We know that the Phillips curve was alive and well during the 1950s through the 1970s, and into the 1980s at the national level. Prices and wages showed significant sensitivity to movements in unemployment during this period. These sensitivities increased when the labour market tightened beyond full employment, indicating a nonlinear relationship. Policymakers allowed the labour market to tighten well beyond full employment levels for a sustained period during the 1960s and, at first, inflation remained low and stable. But several years of tight labour markets resulted in the great inflation of the 1970s.

Since the late 1980s, however, there has been only weak evidence of the sensitivity and nonlinearity of the response of inflation to labour market tightening. Efforts to estimate statistically significant price Phillips curve models using national data have generally failed.

Is the Phillips curve dead?

In a recent paper (Hooper et al. 2019), we argue that there are three reasons why the evidence for a dead Phillips curve is weak.”

Continue reading here.